In an scientific level, the steppe theory proposed by Gimbutas and thereafter reloaded by David Anthony is dead (the kurgan theory 2.0 was about Yamnayans delivering the Germainc languages to the CW culture through the Usatovo culture by a kind of cultural fashion).
The many red and orange alarms were never heard much, but now the final problem was settled by ancient genetics: a sure IE culture as the CW, from where it could split Germanic and Balto-Slavic, had a different Y-DNA than that found in Yamnaya (R1b-Z2105); moreover the steppe cultures possibly linked to the Indo-Iranian branch (Andronovo, Sintashta, Srubnaya) appear to be CW copies in the autosomal and Y-DNA level. The followers of the steppe theory had by some time the hope to find out the CW Y-DNA in the west steppes, along with the right Y-DNA of the Bell Beakers in West Europe, and all speaking the same language ...
Now with the new paper "The genomic history of southeastern Europe" the hope to find and confirm such hopes and theories vanishes. I think it is quite hard to read there:
and
At the Y-DNA level the paper confirms that Yamna was R1b-Z2013 (along ancestor clades) in Samara (x7), Kalmukya (x4), and derived cultures of Staliningrad, Vucedol and Poltavka (x5). The case about Ukranian HGs speaking a form of proto-IE to become Yamnayans is not tenable: the paper states that Ukranian HGs had diverse origins (in Vasilevka there were R1b1a2, I2a1, R1a, I2a2) and had a 31% of WHG in autosomals, such 31% lowered to 1% in Yamnayans, so that only a strong colonization from elsewhere could explain this: even if Yamnayans were true IE speakers, the previous HG population could not be so... and then IE comes from elsewhere. Even in Samara the change is quite obvious: the HG was R1b1a1a (the most eastern R1b), but the Eneolithic samples were Q1a, R1a1 and R1a1a.
The paper suggest to follow the CHG component to Anatolia to find the IE urheimat, but this case is quite difficult to accept: Anatolian imposed over Hatti, Armenian was not local (Urartian placenames have not followed in such language it's own evolutions), and the Caucasus, a catch-all net, is not conserving some unique IE branch (Ossetian is derived from Scythian). The case would be to look further, to the Zagros (as the paper itself also points)... or even in Medieval Russia, if it's the Y-DNA what matters most.
So what is really IE at the end? by now the sure cultures related to such language family display high amounts of R1a and EHG ancestry. The other option left would be to follow the CHG ancestry, its spread, and the clades associated (by now J1, J2). What place would have the R1b-L51 clades is now more difficult to ascertain.